To establish the effect of long-term treatment with raloxifene hydrochloride, compared with placebo, on rate of new vertebral fractures in osteoporotic postmenopausal women with and without prevalent vertebral fractures.